OK gents, started a fresh game, 11 AI opponents against just me, all on a team. I set everything to the most difficult settings, and just rolled in the way I wanted to.
Now it's turn 48, and I am ready to start the invasion. My doom squad is up, and there is nothing the AI will be able to do about it:
The units are effectively unstoppable at this point:
So a couple more turns capturing the back landmarks, and I will wash over them like a wave.
But as per usual, the computer has already thrown it. There are five NPC factions:
And the NPCs are already at war with the AIs:
And the NPC armies are already everywhere:
And as usual, they are wiping out the AI resistance quick, like here where the AI hero stack is about to get totally jacked:
So I expect there won't be anything left when I get there. They are just doing the work too fast for me.
And also as usual, the computer's economic foundation is garbage:
I think that if you just picked sectors at random, you would do better than this. Is it purposefully trying to avoid compatibility? I honestly can't see but maybe one or two fully compatible sectors.
Past that, there are gobs of flaws. Like the psi-fish rewards being stolen when they evolve. So I leveled these hunters up from crysalis:
And naturally the evolution re-sets when the unit fully evolves:
And this is the problem with most units like this. They can't figure out how to make the transition work. Happens with most innate abilities and city bonuses. They always lose like 2/3 of their bonuses when they evolve.
And there are plenty of other errors, such as this city that magically lost a sector expansion because of the Promethian landmark event:
The landmark appeared on the sector that was being annexed, and it just cancelled the annex and blocked the city from expanding elsewhere. Which completely ruined this city. This was initially an excellent place for my capital - it was in reach of 3 gold/silver landmarks and had 2 industry sectors. It would have had it all, but now with only 3 more sectors maximum? DOA. Plus because of where the city is located, it is extremely difficult to get another city half as good to take its place. It almost has to be razed and started again.
It's just tough. And this is how it goes almost every time. The games increasingly all look the same, with few exceptions. Maybe some different details, but this is how it always is once you get to a certain point.
So what is happening right and wrong here? The game does a lot right, but carries a lot wrong with it. Let me hit a few.
Flagship / Paradox rightly simplified city management. The point of management in a game like this is to stack one more level of strategy on top of a complex combat simulator. And that part is great. You are forced to balance the key components: troops (maintenance), tech upgrades, and growth to scale better than your opponents. The winning strategy maximizes growth while keeping enough energy and forces to ensure a commanding lead. You can monkey with that some, such as running an all-in, to punish greedy players.
So simplification went a long way. Further, there are only four resource types in AOW PF, not the whopping 49 resources plus from Civ 6. Add onto that the legion of buildings, units and pointless upgrades in Civ, and you have a handful of viable strategies buried in an ocean of nonsense. Further, the most boring parts of city management are now phased out, such as building roads and setting up trade routs. This is all now instantaneous, and handled fairly well in dialogue screens.
But where everything goes wrong is the same as always. The computer does not know how to prioritize builds, so it relies almost completely on gobs of cheating. According to Akazury, this looks like:
Which still isn't enough. As is well described in other threads, the computer is incompetent. It's incompetent, and it is pretty clear that Flagship / Paradox tried to do it right, but couldn't. So the computer is still stuck on routines that have limited strategic value. In essence, I could summarize the computer's entire combat strategy as: maximize damage on 1 or 2 of the weakest units, inflicting highest current-turn damage on them, even if means hopelessly over-extending. Then die miserably.
So what you will see is armies rushing headlong into entrenched killing-field style concaves that just wipe them out, maybe killing 1 or possibly 2 minor units. Check it out yourself. Attack a big mob of computer units in a city sometime. It will reliably run out of the city walls through a tight choke (city ramps), ensuring that it will be hit with maximum AOE attacks, plus exposing its long-range attackers to all sorts of damage.
These are all problems of one kind. The result is an AI too limited to make any true strategy work. You can only get real strategy if the computer is capable of generating it. It is clear that is not happening yet. So stop making it try. If you want a satisfying PVE experience, you CAN'T DO THIS.
Try these for now:
1. Stop it with the NPC factions. The computer cannot handle them, and please stop making it try. Take all race relations away from the computer and make it dependent on the player. Assume that the computer always gets the quest reward if the timer runs out. Plus remove the AI's NPC faction penalties. Then allow the player to steal quest rewards if he gets to them and wins before the timer runs out. Make it so that the computer will lose relations, and the player will get relations, if the player can snipe the AI quests. You may have to lower rewards, but anything will help.
2. Stop it with AI city management. Give every city a base level of production, a high one, and don't have the AI build buildings in its cities. Have them basically build all-in-one exploitation buildings that convert when a player conquers the city. Then keep the buildings the player selects, but ignore them, assuming it was still all-in-one if the AI conquers it back.
3. Allow the players real flexibility on start-up. Allow larger starting area scans, and allow the players to specify the number of colonizers it starts with, plus the population in each colonizer. For instance, if I started the AI with 3 12-pop colonizers to my 1, that may help.
4. Finish the damn game. Get the build order que finished for god's sake. Seriously.
5. And stop it with the sector exploitation. Have the exploitation building automatically build INDEPENDENT of the buildings and units in the que. Pick what you want it to be in the end, and have it convert within the resource type (food biofarm to food export) in 1 turn, or convert to other sector exploitations in 2 turns (food biofarm to energy optimization). Don't make me go in and manually unbuild 2 levels and re-build 2 levels from the que.
6. Fix the unit evolution / capture. The new unit combinations are some of the most interesting parts of the game. It would be great if they worked.
7. Stop it with the covert operations. First, allow it to be disabled completely. Then consider a PVE version which works on buffing your own cities and units, but not debuffing the opposing player. Currently, the only thing the ops do is drag the game out senselessly. Maybe doomsday is the only alternative.
8. Have multiple attack routines. I can currently only see one combat routine. For instance, all units with grenades always run in starting on turn 2 and throw them. It doesn't really seem to matter what they can hit, they always do just that. This is so predictable that I basically set one or two tough units up front to soak up the grenades. Then I use mine to much better effect. Have the AI mix up a few styles, and see what works. Hell, maybe have a player-base competition for adjustments to the strategies. Whatever the case, have a few approaches, not just one suicidal one.
9. Allow units to upgrade to the bonuses of a city. Whatever city a unit is in, have them pay an upgrade fee, and upgrade to whatever build upgrades the city offers. Maybe even have a 3 turn timer or some such. It feels bad that the units that saved your life 10 times are cursed to suck their whole lives because the game won't allow them to change clothes.
K then. That was a lot.
Thoughts?
Now it's turn 48, and I am ready to start the invasion. My doom squad is up, and there is nothing the AI will be able to do about it:
The units are effectively unstoppable at this point:
So a couple more turns capturing the back landmarks, and I will wash over them like a wave.
But as per usual, the computer has already thrown it. There are five NPC factions:
And the NPCs are already at war with the AIs:
And the NPC armies are already everywhere:
And as usual, they are wiping out the AI resistance quick, like here where the AI hero stack is about to get totally jacked:
So I expect there won't be anything left when I get there. They are just doing the work too fast for me.
And also as usual, the computer's economic foundation is garbage:
I think that if you just picked sectors at random, you would do better than this. Is it purposefully trying to avoid compatibility? I honestly can't see but maybe one or two fully compatible sectors.
Past that, there are gobs of flaws. Like the psi-fish rewards being stolen when they evolve. So I leveled these hunters up from crysalis:
And naturally the evolution re-sets when the unit fully evolves:
And this is the problem with most units like this. They can't figure out how to make the transition work. Happens with most innate abilities and city bonuses. They always lose like 2/3 of their bonuses when they evolve.
And there are plenty of other errors, such as this city that magically lost a sector expansion because of the Promethian landmark event:
The landmark appeared on the sector that was being annexed, and it just cancelled the annex and blocked the city from expanding elsewhere. Which completely ruined this city. This was initially an excellent place for my capital - it was in reach of 3 gold/silver landmarks and had 2 industry sectors. It would have had it all, but now with only 3 more sectors maximum? DOA. Plus because of where the city is located, it is extremely difficult to get another city half as good to take its place. It almost has to be razed and started again.
It's just tough. And this is how it goes almost every time. The games increasingly all look the same, with few exceptions. Maybe some different details, but this is how it always is once you get to a certain point.
So what is happening right and wrong here? The game does a lot right, but carries a lot wrong with it. Let me hit a few.
Flagship / Paradox rightly simplified city management. The point of management in a game like this is to stack one more level of strategy on top of a complex combat simulator. And that part is great. You are forced to balance the key components: troops (maintenance), tech upgrades, and growth to scale better than your opponents. The winning strategy maximizes growth while keeping enough energy and forces to ensure a commanding lead. You can monkey with that some, such as running an all-in, to punish greedy players.
So simplification went a long way. Further, there are only four resource types in AOW PF, not the whopping 49 resources plus from Civ 6. Add onto that the legion of buildings, units and pointless upgrades in Civ, and you have a handful of viable strategies buried in an ocean of nonsense. Further, the most boring parts of city management are now phased out, such as building roads and setting up trade routs. This is all now instantaneous, and handled fairly well in dialogue screens.
But where everything goes wrong is the same as always. The computer does not know how to prioritize builds, so it relies almost completely on gobs of cheating. According to Akazury, this looks like:
| Difficulty | Production | Food/Energy | Knowledge/Influence/Cosmite | Extra Units | Technology | Cities v. Player* |
|---|---|---|---|---|---|---|
| Easy | +0, x1 | +0, x1 | +0, x1 | None | Aquatic Deployment | +1 |
| Normal | +0, x1 | +0, x1 | +0, x1 | None | Aquatic Deployment | +2 |
| Hard | +0, x1.165 | +10, x1.33 | +5, x1.33 | None | Aquatic Deployment | +3 |
| Very Hard | +0, x1.33 | +20, x1.66 | +10, x1.66 | +1 | Aquatic Deployment | No Limit |
| Extreme | +0, x1.5 | +40, x2 | +20, x2 | +2 | Aquatic Deployment | No Limit |
Which still isn't enough. As is well described in other threads, the computer is incompetent. It's incompetent, and it is pretty clear that Flagship / Paradox tried to do it right, but couldn't. So the computer is still stuck on routines that have limited strategic value. In essence, I could summarize the computer's entire combat strategy as: maximize damage on 1 or 2 of the weakest units, inflicting highest current-turn damage on them, even if means hopelessly over-extending. Then die miserably.
So what you will see is armies rushing headlong into entrenched killing-field style concaves that just wipe them out, maybe killing 1 or possibly 2 minor units. Check it out yourself. Attack a big mob of computer units in a city sometime. It will reliably run out of the city walls through a tight choke (city ramps), ensuring that it will be hit with maximum AOE attacks, plus exposing its long-range attackers to all sorts of damage.
These are all problems of one kind. The result is an AI too limited to make any true strategy work. You can only get real strategy if the computer is capable of generating it. It is clear that is not happening yet. So stop making it try. If you want a satisfying PVE experience, you CAN'T DO THIS.
Try these for now:
1. Stop it with the NPC factions. The computer cannot handle them, and please stop making it try. Take all race relations away from the computer and make it dependent on the player. Assume that the computer always gets the quest reward if the timer runs out. Plus remove the AI's NPC faction penalties. Then allow the player to steal quest rewards if he gets to them and wins before the timer runs out. Make it so that the computer will lose relations, and the player will get relations, if the player can snipe the AI quests. You may have to lower rewards, but anything will help.
2. Stop it with AI city management. Give every city a base level of production, a high one, and don't have the AI build buildings in its cities. Have them basically build all-in-one exploitation buildings that convert when a player conquers the city. Then keep the buildings the player selects, but ignore them, assuming it was still all-in-one if the AI conquers it back.
3. Allow the players real flexibility on start-up. Allow larger starting area scans, and allow the players to specify the number of colonizers it starts with, plus the population in each colonizer. For instance, if I started the AI with 3 12-pop colonizers to my 1, that may help.
4. Finish the damn game. Get the build order que finished for god's sake. Seriously.
5. And stop it with the sector exploitation. Have the exploitation building automatically build INDEPENDENT of the buildings and units in the que. Pick what you want it to be in the end, and have it convert within the resource type (food biofarm to food export) in 1 turn, or convert to other sector exploitations in 2 turns (food biofarm to energy optimization). Don't make me go in and manually unbuild 2 levels and re-build 2 levels from the que.
6. Fix the unit evolution / capture. The new unit combinations are some of the most interesting parts of the game. It would be great if they worked.
7. Stop it with the covert operations. First, allow it to be disabled completely. Then consider a PVE version which works on buffing your own cities and units, but not debuffing the opposing player. Currently, the only thing the ops do is drag the game out senselessly. Maybe doomsday is the only alternative.
8. Have multiple attack routines. I can currently only see one combat routine. For instance, all units with grenades always run in starting on turn 2 and throw them. It doesn't really seem to matter what they can hit, they always do just that. This is so predictable that I basically set one or two tough units up front to soak up the grenades. Then I use mine to much better effect. Have the AI mix up a few styles, and see what works. Hell, maybe have a player-base competition for adjustments to the strategies. Whatever the case, have a few approaches, not just one suicidal one.
9. Allow units to upgrade to the bonuses of a city. Whatever city a unit is in, have them pay an upgrade fee, and upgrade to whatever build upgrades the city offers. Maybe even have a 3 turn timer or some such. It feels bad that the units that saved your life 10 times are cursed to suck their whole lives because the game won't allow them to change clothes.
K then. That was a lot.
Thoughts?
Attachments
- 2
- 1
